46 GENERAL DISCUSSION GENERAL DISCUSSION Dr. P. Flodin (Inst. Biochemistry, Upsala) said: During work on the elimina- tion of adsorption in fdter paper performed at the Institute of Biochemistry, Upsala, attempts have been made to give the surface of the paper the same charge as the substances to be investigated. By repulsion the substances would then be prevented from coming into contact with the surface at which secondary forces may cause adsorption. Hoffpauir and Guthrie 1 have treated cotton with 2-aminoethylsulphuric acid to give it anion-exchanging properties. When filter paper was treated in the same manner it was found to keep its wet strength and its water-absorptive property and was generally suitable for filter paper electrophoresis in the same way as ordinary paper.The presence of free amino groups was demonstrated by its strong reaction with ninhydrin. To make the paper suitable for experiments with positively charged substances it had to be converted to its -NH3+ form with e.g. hydrochloric acid. Positively charged proteins, e.g. serum albumin, P-lactoglobulin and legumin at low pH, have been run without detectable adsorption although these proteins are strongly adsorbed on untreated paper. Surprisingly it was found that even negatively charged proteins could be run without irreversible adsorption. Blood serum has been separated into components at pH 8-6 in a few experiments on pre-treated paper. The possibilities of this modified paper for use in zone electrophoresis and chromatography will be further investigated.Dr. R. Consden (Canadian Red Cross Memorial Hospital, Maidenhead) said : A simple paper electrophoresis apparatus, which is in regular use for routine investigations of pathological sera and for ionophoresis of amino-acids, peptides and carbohydrates, consists of a rectangular glass tray 35 x 20 cm* and 5 cm in depth, in which is placed a close-fitting rectangular glass frame. This frame is constructed of two rectangles of glass rod mounted one over the other, and 1 Hoffpauir and Guthrie, J . Biol. Chern., 1949, 178, 207.GENERAL DISCUSSION 47 separated from each other by about 1.5 cm by sealed-on pieces of glass, or by pieces of cork. The frame stands on short legs, and it carries the paper sheet and keeps it stretched horizontally. The free ends of the paper pass over the edges of the tray and dip into Perspex boxes containing electrolyte and electrodes.After the paper is wetted with electrolyte, and the solution to be analyzed is applied, an organic liquid (e.g. chlorobenzene) is poured into the tray until the paper is covered. Glass pieces are placed in the electrolyte boxes to keep the paper ends away from the electrodes, and the whole tray may be covered with a glass lid. Electrophoresis may be carried out either across the shorter width of the tray or across the longer. Voltage gradients are 10-1 1 V/cm in the former method, and 6-7 V/cm in the latter. The organic liquid prevents evaporation of electrolyte from the paper, and acts as a cooler. The frame is convenient for dealing with wet paper, after the electrophoresis, and eliminates undesirable contamination by contact with the hands, when chromatography is to be subsequently carried out.For a given voltage gradient and temperature, movement is linear with time, and good repro- ducibility for a given substance is obtained. Dr. E. L. Durrum (Army Medical Service Grad. School, Washington, D.C.) said : Professor Tiselius has provided a very complete list of references. However, a paper by v. Klobusitzky and Konig 1 has recently come to our attention in which a paper strip electrophoresis technique developed in 1937 by Konig was used in 1939 to separate a yellow pigment fraction in snake venom. It is interesting to note that this work was done before paper chromatography became so widely used. Also, in connection with the work of Svensson and Brattsten and the work of Grassmann and Hannig on continuous electrophoresis it may be in order to call attention to work of Haugaard and Kroner, who in 1948 applied for a United States patent 2 which has been issued only recently and which clearly discloses the principle of this type of separation independently arrived at by these workers.I should like to comment on one other point which Prof. Tiselius has mentioned. That is the difficulties due to tailing caused by adsorption phenomena. Except in a few specific cases, in our own experience with the free hanging strip technique, tailing attributable to adsorption is less commonly encountered than that type due to paper overloading. Yet another type of tailing appears to be related to the following : when paper is supported on one or both sides, even when the supporting surface is siliconed, a thin film of electrolyte lies at the interface between the paper and supporting medium in which convection may not be efficiently prevented and in which film the mobility may be different from that in the body of the paper.In our opinion, this is often the reason for the so-called tailing encountered in closed strip techniques. On the other hand, when the strip hangs free, this con- dition is not present, and we believe this to be the explanation why in general zones are more sharply defined in this variation. Of course, if it is desired to measure mobilities, the free-strip technique (where evaporation is permitted) can be used with confidence only if reference substances of known mobility and similar adsorption properties are studied simultaneously, and it is probable that even under the best conditions measurements of this type are not so satisfactorily made as in the method where the strip is totally enclosed and evaporation thus prevented, using, of course, corrections for electroendosmosis and those for migration path length proposed by Kunkel and Tiselius.Dr. D. L. Mould (Rowett Research Institute) said: Prof. Tiselius has men- tioned the application of electro-osmosis for " zone ultrafiltration ". A micro- method has been developed for the fractionation of uncharged molecules such as enzyme-synthesized polysaccharide preparations.3 Electro-osmotic streaming 1 von Klobusitzky and Konig, Arch. exper.Path. Pharmakol, 1939, 192, 271. 2 U.S. Pat. 2,555,487 issued to G. Haugaard and T. D. Kroner, filed on 27th February, 3 Mould and Synge, Biochem. J., 1951, 50, 11 ; Analyst, 1952 (in press). 1948.48 GENERAL DISCUSSION forces a solution through a strip of collodion ultrafiltration membrane and ad- sorption and the molecular sieve effect of the porous membrane structure both tend to retard the larger molecules. Migration zones are formed related to the DP of the synthesized polysaccharide. Dr. E. Barbu (Paris) said: We are interested in the possibility of paper zone electrophoresis giving some indication of the size and shape polydispersity of protein aggregates. Protein aggregates of different sizes can show the same mobilities in boundary electrophoresis but by paper zone electrophoresis we can differentiate the larger aggregates which migrate more slowly than the smaller.An example is shown in fig. 1, 2 and 3 of the paper by Barbu and Joly. (For further details, see Barbu, Macheboeuf and Rebeyrotte, Bull. Soc. Chim. Biol., 1952 (in press)). Dr. L. Robert (Facufte' de Medecine de Paris) said : The Maillard reaction may be used with advantage for the qualitative and quantitative estimation of proteins or amino acids in paper electrophoresis. On exposing the dried paper strips to acetaldehyde vapour in a closed system, the well-known reddish-brown, highly fluorescent, pigment is formed and this may be eluted and quantitatively estim- ated by fluorimetry. This method can be made more sensitive by treating the paper strip with a dilute solution of morpholine, before exposing it to acetaldehyde.The method is carried out in our laboratory as follows : (i) after electrophoretic migration, the filter paper strips are dried, followed by (ii) immersion in a 4 % solution of morpholin in methylal, for 5 min, (iii) immersion in pure methylal to remove the excess of morpholine for 4 min, (iv) immersion in ether for 4 min, (v) drying in air, (vi) exposure to acetaldehyde vapours. For this purpose a 20-30 % solution of pure acetaldehyde in methylal is recommended in a wide Petri dish. The paper strips are placed on glass frames and covered with glass jars. About 15-20 min exposure to acetaldehyde vapours is usually sufficient to give a fairly intense pigment formation. If not sufficiently rinsed, the paper turns brown, because morpholine alone gives with acetaldehyde (in absence of proteins) a reddish-brown pigment.Dioxane, however, removes this pigment without extracting the protein pigment. It may be mentioned that paper strips when treated with the acetaldehyde+ morpholine pigment (prepared by distilling acetaldehyde into a morpholine + ether solution when the pigment separates out and may be used in a butanol solution) a somewhat lighter coloration of the protein is obtained. This modification has the advantage that the pigment bound by the protein is much more easily eluted than the pigment formed with the protein. Elution may in fact sometimes present a problem which may be overcome only by a careful choice of solvents. Pigment formed with aminoacids, casein, and some other proteins are readily soluble in butanol t- acetic acid + water (2/1/2) mixture at 100' C or in butanol + piperidine + water (2/1.5/2).Form- amide is also a suitable solvent, Serum-protein pigments are much more difficult to obtain in solution. We are trying a special fluorimeter based on the same principles as the ordinary densitometers used in paper electrophoresis. Quanti- tative data and other details will be published elsewhere. It should be mentioned that this principle may also be applied to paper chromatography. The use of fluorescence in the quantitative estimation may possibly widen the present con- centration limits in paper electrophoresis. Dr. A. G. Ogston (Oxford University) said: Dr. Weber has attempted to ex- plain the lack of effect of dilution upon the apparent degree of dissociation of albumin by assumptions about the kinetic orders of the association and dissoci- ation reactions.Whatever such assumptions are made do not affect the require-GENERAL DISCUSSION 49 ment that a dissociation equilibrium must obey thermodynamic laws. These demand that, if dilution does not affect the equilibrium, the reaction does not affect the total number of solute particles. It would, therefore, be necessary to assume that some other substance, which is also diluted as the albumin is diluted, takes part in the dissociation reaction; for examples, if A2, A are the associated and dissociated a1 bumin, A~ + BJAB + A or A2 -r B2 7 2AB. Dr. K. 0. Pedersen (Lrpsala) said: I wonder if bovine serum albumin really dissociates in acid solutions under the conditions given by Dr.Weber in his paper. In some studies I made some years ago on the sedimentation, diffusion and viscosity of bovine serum albumin in acid solutions I found that a pronounced change in the protein molecule takes place in the region pH 2.3 to pH 3 (temperature 10" to 35" C). Contrary to my expectation I had to assume that the molecular weight of the serum albumin remained constant. I had to explain the decrease in the sedimentation and diffusion coefficients, and the increase in the relative viscosity by assuming a partly reversible unfolding of the molecule. Increase of the ionic strength of the solution from 0.1 to 0.2, or from 0.2 to 0-3, diminished the effect thus indicating the presence of electrostatic charge effects even in solutions having an ionic strength of 0.1 and 0.2.I think one must be very careful in drawing conclusions from single ultra- centrifugal, diffusion or osmotic pressure experiments in this pH region without knowing more about the magnitude of the electrostatic charge effect. Dr. B. Robert (Institut Pasteur, Paris) said: Barbu and Macheboeuf have demonstrated that most proteins change their native configuration at pH values above 9. Their physical and chemical properties alter instantaneously in alkaline media. We were able to show 1 that this configurational modification rendered accessible the -SH groups of serum albumin, which could reduce ferricyanide at pH 8.5 and were oxidizable by air above pH 10. Kinetic analysis of the oxidation of alkaline protein solutions allow some conclusions to be drawn concerning the mechanism of the structural change occurring in proteins at these high pH values.We found, for example, that the dissociation of a proton (with a pK value of about 13) is essential for autoxidation. We concluded that guanidine groups of arginine and cystine S-S bridges are concerned in maintaining the stability of serum albumin molecules in agreement with the finding of Klotz et al. Structural changes occurring above pH 9 are irreversible because of the above- mentioned oxidation of the molecule. It is therefore possible to obtain this struc- tural change in a reversible manner when such side reactions are avoided. Dr. G. S. Adair (Low Temperature Statiorz, Cambridge) said : Measurements of solubilities and specific properties described by Falconer, Jenden and Taylor give one of the most sensitive tests for the purity of a protein.I should like to amplify one point. In experiments made with varying amounts of protein and a constant solvent, two procedures are possible. (i) The molality or the ratio of salt to water can be kept constant. (ii) The activity or effective concentration of the salt may be kept constant. A close approximation to the second procedure might be obtained by dialyzing a mixture of crystals and mother liquor against a salt solution, and using mixtures containing crystals and dialysate for solubility tests. The difference between methods (i) and (ii) may be insignificant in experiments made with small quantities of protein, as used in specific property tests. The second method may prove useful for work on more concentrated solutions of proteins. Dr. L. Nanninga (Leiden) said: From fig. 1, curves I and 11, of Falconer, Jenden and Taylor's paper, it follows that Se/Si increases when cprot. decreases. 1 Robert and Macheboeuf, Bull. Sur. Citeni. Biol., 1953, 35 (in press).50 GENERAL DISCUSSION However, I do not agree that it follows that ki > ke, since the values j3i and in the solubility formula are not accounted for. As I remains constant with varying cprot. there can be no differentiation with respect to 1 here. Dr . D . B . Taylor (California University) said : The differentiation referred to by Dr. Nanninga was of the equation relating the ratio of the solubilities of the enzyme and of the impurity to the ionic strength. These solubilities are ex- ponential functions of ionic strength in accordance with Cohn’s equation. The differentiation would therefore appear to be in order. The /3 constants do not appear in the equation since we are concerned with the rates of salting out of the proteins as a function of salt strength.